(NTHi) Infection and Airway Inflammation in COPD - Semantic Scholar

RESEARCH ARTICLE

Relationships between Mucosal Antibodies, Non-Typeable Haemophilus influenzae (NTHi) Infection and Airway Inflammation in COPD Karl J. Staples1,2*, Stephen Taylor3, Steve Thomas3, Stephanie Leung3, Karen Cox1, Thierry G. Pascal4, Kristoffer Ostridge5, Lindsay Welch5, Andrew C. Tuck1, Stuart C. Clarke1,2,5, Andrew Gorringe3, Tom M. A. Wilkinson1,2,5

a11111

1 Clinical & Experimental Sciences, University of Southampton Faculty of Medicine, Southampton General Hospital, Tremona Road, Southampton, United Kingdom, 2 Wessex Investigational Sciences Hub, University of Southampton Faculty of Medicine, Southampton General Hospital, Tremona Road, Southampton, United Kingdom, 3 Public Health England, Porton Down, Salisbury, United Kingdom, 4 GSK Vaccines, Wavre, Belgium, 5 Southampton NIHR Respiratory Biomedical Research Unit, Southampton General Hospital, Tremona Road, Southampton, United Kingdom * [email protected]

OPEN ACCESS Citation: Staples KJ, Taylor S, Thomas S, Leung S, Cox K, Pascal TG, et al. (2016) Relationships between Mucosal Antibodies, Non-Typeable Haemophilus influenzae (NTHi) Infection and Airway Inflammation in COPD. PLoS ONE 11(11): e0167250. doi:10.1371/journal.pone.0167250 Editor: John S Tregoning, Imperial College London, UNITED KINGDOM Received: September 8, 2016 Accepted: November 10, 2016 Published: November 29, 2016 Copyright: © 2016 Staples et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: This work was supported by GlaxoSmithKline Biologicals, SA, Belgium via a Collaborative Research & Development Agreement (CRADA). The funder provided support in the form of salaries for TGP and project management but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of TGP are articulated in the ‘author contributions’

Abstract Non-typeable Haemophilus influenzae (NTHi) is a key pathogen in COPD, being associated with airway inflammation and risk of exacerbation. Why some patients are susceptible to colonisation is not understood. We hypothesised that this susceptibility may be due to a deficiency in mucosal humoral immunity. The aim of our study (NCT01701869) was to quantify the amount and specificity of antibodies against NTHi in the lungs and the associated risk of infection and inflammation in health and COPD. Phlebotomy, sputum induction and bronchoscopy were performed on 24 mild-to-moderate COPD patients and 8 age and smokingmatched controls. BAL (Bronchoalveolar lavage) total IgG1, IgG2, IgG3, IgM and IgA concentrations were significantly increased in COPD patients compared to controls. NTHi was detected in the lungs of 7 of the COPD patients (NTHi+ve—29%) and these patients had a higher median number of previous exacerbations than NTHi-ve patients as well as evidence of increased systemic inflammation. When comparing NTHi+ve versus NTHi-ve patients we observed a decrease in the amount of both total IgG1 (p = 0.0068) and NTHi-specific IgG1 (p = 0.0433) in the BAL of NTHi+ve patients, but no differences in total IgA or IgM. We observed no evidence of decreased IgG1 in the serum of NTHi+ve patients, suggesting this phenomenon is restricted to the airway. Furthermore, the NTHi+ve patients had significantly greater levels of IL-1β (p = 0.0003), in BAL than NTHi-ve COPD patients.This study indicates that the presence of NTHi is associated with reduced levels and function of IgG1 in the airway of NTHi-colonised COPD patients. This decrease in total and NTHI-specific IgG1 was associated with greater systemic and airway inflammation and a history of more frequent exacerbations and may explain the susceptibility of some COPD patients to the impacts of NTHi.

PLOS ONE | DOI:10.1371/journal.pone.0167250 November 29, 2016

1 / 17

Reduced IgG1 in COPD

section. No restrictions were placed on authors regarding the statements made in the manuscript. Competing Interests: TGP is an employee of the GSK group of companies; he owns shares in GSK. These interests do not alter our adherence to PLOS ONE policies on sharing data and materials. The remaining authors have declared that no competing interests exist.

Introduction Non-typeable Haemophilus influenzae (NTHi) is a non-capsulated bacterium that can cause respiratory tract infections throughout life[1]. NTHi can also chronically infect the airways of patients with COPD and is associated with greater airway inflammation[2–4] and with acute exacerbations[2, 5]. Microbiological studies of the COPD airway have identified that the presence of NTHi is associated with loss of microbial diversity[6] and that dynamic changes in NTHi strain carriage contribute to risk of exacerbation[5]. Whilst there is growing evidence from a number of sources for an impaired immune response in the COPD lung[7–9] the mechanisms which lead to susceptibility to NTHi infection are not yet understood. When comparing COPD patients to healthy controls, total serum immunoglobulin (Ig) levels are not reduced, suggesting that it is not a deficiency in overall systemic humoral immunity that underlies risk of airway infection[10]. In fact, the evidence points to certain antibodies (IgD) being elevated in the serum of COPD patients[10]. Previous work has found NTHi infection of the COPD lung despite detection of specific IgG and IgA antibodies in both serum and sputum[11]. However, a decrease in the amount of secretory (s)IgA in the bronchoalveolar lavage (BAL) of COPD patients, has been recently reported in advanced disease[7]. In contrast there have also been reports of increased IgA in COPD lung tissue[12], but to date no study has comprehensively analysed immunoglobulin levels in the human airway in the context of NTHi infection, or in earlier stages of disease. Indeed the immune correlates of protection against this un-encapsulated organism are not fully understood, but murine models of COPD have demonstrated impaired immunoglobulin class switching in disease with associated effects on NTHi clearance[13]. Whether this is a relevant mechanism in man is not understood. The aim of this study was to explore the differences in airway immunoglobulin levels and associated functional competence of mucosal antibody-mediated immunity to NTHi in health and COPD. We further wished to assess how these aspects of mucosal immunity were related to airway infection and inflammation in stable disease.

Materials and Methods Ethics All subjects gave written informed consent and the study (ClinicalTrials.gov: NCT01701869) was approved by the National Research Ethics Service (NRES) Southampton B Committee (12/SC/0304).

Subjects Twenty-four subjects with stable mild and moderate COPD [14] were recruited. COPD diagnosis was confirmed by post-bronchodilator spirometry with a FEV1/FVC ratio of 80% at Q30 and were therefore deemed successful.

Multi Locus Sequence Typing The multi locus sequence typing (MLST) schema for H. influenzae was downloaded from SRST2[18, 19]. SRST2 was used to perform in-silico MLST analysis on all isolates, all resulting in coverage >30.

In-silico PCR for identification of NTHi iga gene Previously published primers were used in iPCRess for in-silico PCR analysis to ascertain presence of iga responsible for iga protease translation in NTHi[20]. Forward primer igaBF1 TGAATAACGAGGGGCAATATAAC and reverse primer igaBR1 TCACCGCACTTAATC ACTGAAT [21]. The mismatch setting in iPCRess was set to 3.

Mapping to reference sequences for the NTHi iga and igaB protease genes Reference sequences for the conserved beta core sequence of iga and full CDS of igaB, genes responsible for iga protease production in H. influenzae were obtained from GenBank (iga– GenBank accession no M87492, bases 4124–4978 igaB GenBank accession number KC607498.1). Isolates were mapped against these reference sequences using SRST2 with settings of minimum coverage at 60 and invoking the ‘report_all_consensus_alleles’ option [19].

Serum and BAL supernatant analyses Total immunoglobulin isotype concentrations were quantified using the MSD platform according to the manufacturer’s instructions (MSD, Rockville, USA). Secretory IgA was quantified using ELISA according to the manufacturer’s instructions (Demeditec Diagnostics GmbH, Kiel, Germany). Cytokine and MMP concentrations in BAL were quantified using a Luminex multiplex immunoassay (R&D systems, Abingdon, UK). Samples were analysed on the Luminex 200 platform (Biorad Bioplex 200, Hemel Hempstead, UK), as per manufacturer’s instructions. Cytokine analysis was performed for IL-1β, IL-2, IL-6, IL-8, IL-10, GM-CSF, IFNγ and TNFα. Matrix metalloprotease (MMP) analysis was performed for MMP-1, -2, -3, -7, -8, -9, -10, -12, -13 and the ECM metalloprotease inducer (EMMPRIN).

Assessment of NTHi-specific antibodies by immunoglobulin binding assay (IBA) Binding of antibodies to NTHi was determined by flow cytometry. NTHi 3224A (sequence type (ST) 259—obtained from GSK Vaccines) was plated onto chocolate agar, supplemented with PolyVitex (Biomerieux SA, France), and incubated at 37˚C with 5% CO2 for 16–24 h. Bacteria were resuspended in 2.5 ml brain heart infusion broth (Oxoid) supplemented with haemin (Sigma) and nicotinamide adenine dinucleotide (Sigma). This was incubated with shaking for 2–3h at 37˚C until the OD620nm was 0.35–0.45. Two microliters of heat


4 / 17


inactivated sera (56˚C for 30 min) was added to appropriate wells of a standard U-bottom 96-well microtitre plate. 198 μl of bacteria, at OD620nm 0.1, in blocking buffer (IBA-BB) (2% BSA in 1x PBS w/v) was added to every well and the plate incubated at 25˚C for 30 min with shaking at 900 rpm. The plate was centrifuged at 3060g for 5 min, each well washed with 200 μl IBA-BB, centrifuged again and the pellets re-suspended in 200μl fluorescent detection antibody conjugated with fluorescein isothiocyanate (FITC) or 200 μl unconjugated secondary antibody, diluted 1:500 in IBA-BB. The conjugated detection antibodies used were either goat anti-human FITC or goat anti-mouse FITC, depending on whether the assay used test sera (human) or test sera and a secondary antibody (mouse). Where the assay was detecting just the test antibody the plate was incubated for 20 min at 4˚C then centrifuged at 3060 g for 5 min, washed twice with 200 μl IBA-BB, and finally centrifuged again and the pellets re-suspended in 200 μl IBA-BB. Where the assay was detecting a secondary antibody the plate was incubated for a further 20 min at 4˚C, after the first of the two wash steps mentioned previously, before being washed twice with IBA-BB. The plate was stored in the dark at 4˚C until analysed by flow cytometry. All tests were performed in duplicate and the following background controls were used in the assay: bacteria only and bacteria plus conjugate only.

Oxidative burst assay (OBA) Bacteria were prepared at 5.0x109/ml in blocking buffer (OBA-BB) (2% Marvel in HBSS+Ca2+ +Mg2+). 5 μl heat inactivated sera were added to appropriate wells of a standard U-bottom 96-well plate. 15 μl of OBA-BB was added to wells containing sera, or an appropriate amount to control wells to give final volume of 40 μl prior to addition of HL60 cells. 10 μl bacteria was added to every well except “cells only” control. The plate was incubated for 15 min at 37˚C with shaking at 900 rpm. 10 μl IgG-depleted human plasma (diluted 1:10 in OBA-BB) was added to appropriate wells. This was then incubated for 7.5min at 37˚C with shaking at 900 rpm. Differentiated HL60 cells at 2.5x107/ml were prepared by centrifuging at 400 g for 5min and then re-suspending in OBA-BB. 25 μl prepared cells were added to all wells along with 25 μl 25 μg/ml Dihydrorhodamine 123 (DHR 123 –Life Technologies, D23806), and incubated for 15 min at 37˚C with shaking at 900 rpm. Microtitre plates were immediately placed on ice and assay fixed with 80 μl 1% formaldehyde in DPBS+0.02% (w/v) EDTA and incubated for 30 min at RT covered with foil, to eliminate light. Wells were analysed immediately on the flow cytometer.

Bacterial flow cytometry Assays were analysed using a Beckman Coulter Cyan flow cytometer equipped with a Cytek 96-well microtitre plate reader. Protocols were initially set-up to analyse profiles of events identified on the cytometer by the forward scatter (FS), measuring the size of the cell, and side scatter (SS), measuring the granularity and internal structural complexity. An analysis ‘gate’ was drawn around the population of interest and a relevant histogram plot created to analyse the fluorescence given off by the event population. For each sample, 10,000 individual events were analysed for fluorescence and a horizontal gate was drawn to include ~10% of the control sample population (bacteria plus conjugate for IBA and cells plus bacteria plus complement for OBA). A Fluorescence Index (FI) was calculated for each sample, which involved the multiplication of the % of events moving into the horizontal gate (%-gated), by the average fluorescence of that population (X-mean). The final result for each test was expressed as the average FI (average FI taken for duplicate test samples) of the test serum sample minus the average FI of the bacteria plus conjugate only control for IBA, expressed as FI-Conj and cells plus bacteria plus complement control for OBA, expressed as FI-C’.


5 / 17


Statistics Analysis of three groups was performed using a Kruskal-Wallis ANOVA followed by a Dunn’s post hoc test. Analysis of two groups was performed using a one-tailed Man-Whitney U test for analysis of previous exacerbations and NTHi-specific Ig when a previous Kruskal-Wallis test had already indicated a difference in total Ig levels. Fishers exact test was used for categorical data (GraphPad Prism v6, GraphPad Software Inc., San Diego, USA). Each subject had two lobes sampled and the mean concentrations of the separate aliquots from each lobe were used. Associations between Ig and spirometry parameters were assessed using Spearman’s correlation with rho and p values presented. Results were considered significant if p 6 months. Blood data shown represents 8 controls, 15 NTHi-ve and 7 NTHi+ve COPD volunteers. BAL data shown represents 8 controls and 14 NTHI-ve and 5 NTHI+ve COPD volunteers. Data were analysed using a Kruskal-Wallis ANOVA with Dunn’s post hoc comparison. * p